/replicate

Replicate Website

Entry point for website replication workflow.

Usage

/replicate <url> [options]

Options:

• --quick - Skip responsive captures, focus on main viewport only
• --animations-only - Focus only on scroll animation capture
• --static-only - Skip animation detection, capture static structure only

Workflow Start

1. Load the skill: skills/website-replicator/SKILL.md
2. Create output directory structure
3. Begin Phase 1 (Initial Load)

Output Location

All captures and generated code go to: ./website-replication-{timestamp}/

What Happens

Phase 1: Navigate & screenshot overview → You may send to Gemini Phase 2: DOM reconnaissance → Build watch list
Phase 3: Animation capture → Monitor scroll changes Phase 4: CSS extraction → Capture styles Phase 5: Asset collection → List images/fonts Phase 6: Analysis & generation → Produce code

Interruption Points

The workflow will pause and ask for input when:

• Screenshot ready for Gemini analysis
• Watch list needs confirmation
• Automated detection fails
• Verification needed

Resume

If interrupted, the workflow state is preserved. Run /replicate --resume to continue.

Begin

Load the target URL and start capture.